Thread: aaagh... postgres is segfaulting?

(apology if this is a duplicate mail client was acting up)

I'm trying a fresh install of postgres on a redhat machine (gcc 2.96) compiled
from source.

I am getting messages along these lines in the logfile, and php is telling me
that the backend abruptly ended the connection.

How can i go about fixing this?

Josh.
LOG:  database system was shut down at 2003-04-09 06:20:55 EDT
LOG:  checkpoint record is at 0/1816C2C
LOG:  redo record is at 0/1816C2C; undo record is at 0/0; shutdown TRUE
LOG:  next transaction id: 5882; next oid: 164471
LOG:  database system is ready
LOG:  server process (pid 29227) was terminated by signal 11
LOG:  terminating any other active server processes
LOG:  all server processes terminated; reinitializing shared memory and
semaphores
LOG:  database system was interrupted at 2003-04-09 06:22:07 EDT
LOG:  checkpoint record is at 0/1816C2C
LOG:  redo record is at 0/1816C2C; undo record is at 0/0; shutdown TRUE
LOG:  next transaction id: 5882; next oid: 164471
LOG:  database system was not properly shut down; automatic recovery in
progress
LOG:  redo starts at 0/1816C6C
LOG:  ReadRecord: record with zero length at 0/181F160
LOG:  redo done at 0/181F13C
LOG:  database system is ready
LOG:  server process (pid 18767) was terminated by signal 11
LOG:  terminating any other active server processes
WARNING:  Message from PostgreSQL backend:
        The Postmaster has informed me that some other backend
        died abnormally and possibly corrupted shared memory.
        I have rolled back the current transaction and am
        going to terminate your database system connection and exit.
        Please reconnect to the database system and repeat your query.
WARNING:  Message from PostgreSQL backend:
        The Postmaster has informed me that some other backend
        died abnormally and possibly corrupted shared memory.
        I have rolled back the current transaction and am
        going to terminate your database system connection and exit.
        Please reconnect to the database system and repeat your query.
LOG:  all server processes terminated; reinitializing shared memory and
semaphores
LOG:  database system was interrupted at 2003-04-09 06:27:20 EDT
LOG:  checkpoint record is at 0/184954C
LOG:  redo record is at 0/184954C; undo record is at 0/0; shutdown FALSE
LOG:  next transaction id: 6997; next oid: 172663
LOG:  database system was not properly shut down; automatic recovery in
progress
LOG:  redo starts at 0/184958C
LOG:  ReadRecord: record with zero length at 0/186DF6C
LOG:  redo done at 0/186DF48
LOG:  database system is ready

On Wednesday 09 Apr 2003 12:01 pm, Joshua Moore-Oliva wrote:
> (apology if this is a duplicate mail client was acting up)
>
> I'm trying a fresh install of postgres on a redhat machine (gcc 2.96)
> compiled from source.

So RedHat 7.3?

> I am getting messages along these lines in the logfile, and php is telling
> me that the backend abruptly ended the connection.

I take it you also get crashes from the command-line as well as PHP?

> How can i go about fixing this?
>
> Josh.
> LOG:  database system was shut down at 2003-04-09 06:20:55 EDT
> LOG:  checkpoint record is at 0/1816C2C
> LOG:  redo record is at 0/1816C2C; undo record is at 0/0; shutdown TRUE
> LOG:  next transaction id: 5882; next oid: 164471
> LOG:  database system is ready
> LOG:  server process (pid 29227) was terminated by signal 11
> LOG:  terminating any other active server processes
> LOG:  all server processes terminated; reinitializing shared memory and
> semaphores
> LOG:  database system was interrupted at 2003-04-09 06:22:07 EDT

A sig11 is frequently caused by bad memory or some other hardware fault.
Probably worth running a memory tester for a few hours to check. These
problems can appear only in key apps and are a pig to track down.

Also, make sure you've got the latest gcc rpm's from RedHat. I've got a vague
memory that they shipped a pre-release version initially.

Finally, give the rpm version a quick go if you can. I've not noticed any such
problems on RH7.3 with the current rpms.

If you suspect an actual fault in the code, the developers will be wanting a
backtrace from gdb (on the backend). Check the archives/google for details of
what info is useful to provide.

--
  Richard Huxton